SEIT: Structural Enhancement for Unsupervised Image Translation in Frequency Domain

Authors: Zhifeng Zhu, Yaochen Li, Yifan Li, Jinhuo Yang, Peijun Chen, Yuehu Liu

AAAI 2024 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental The extensive experimental results well demonstrate the effectiveness of the proposed method.
Researcher Affiliation Academia 1School of Software Engineering, Xi an Jiaotong University 2Institute of Artificial Intelligence and Robotics, Xi an Jiaotong University z1965761380@stu.xjtu.edu.cn, yaochenli@mail.xjtu.edu.cn, {3121358033, jinhuo, 3123358029}@stu.xjtu.edu.cn, liuyh@mail.xjtu.edu.cn
Pseudocode No The paper does not contain structured pseudocode or algorithm blocks.
Open Source Code No The paper does not provide an explicit statement or link for open-source code availability for the described methodology.
Open Datasets Yes The datasets we use include SYNTHIA(Ros et al. 2016), GTA5(Richter et al. 2016), Cityscapes(Cordts et al. 2015) and BDD(Yu et al. 2020).
Dataset Splits No The paper does not provide specific train/validation/test dataset splits, percentages, or explicit methodology for data partitioning.
Hardware Specification Yes All experiments are conducted on a single RTX 3090 GPU.
Software Dependencies No The paper mentions software components like 'Adam optimizer' and 'VGG network' but does not provide specific version numbers for any programming languages, libraries, or frameworks used for implementation.
Experiment Setup Yes The batch size is set to 1. We use the Adam optimizer with β1 = 0.5 and β2 = 0.999. The initial learning rate is set to 0.0002 and the step decay learning strategy is used, with the learning rate decaying to half of the original learning rate every 5 epochs. The model is trained for 100 epochs. Following previous work, the loss weight in equation 14 is set to 1.0, 2.0, and 1.0, respectively.